Youtube daily report Sep 8 2018

-I want to do something fun with you if you don't mind.

There are a lot of these really popular quizzes online

on sites like BuzzFeed where you answer a series of questions

and they tell you which Jonas brother you are.

[ Laughter ]

Have you seen these? Have you heard of these?

-I've seen these. -Yeah. Well --

-Never done it. This is strange.

-We have one right here. -Oh, boy.

-And I was just wondering if you wanted to do it

and see which Jonas brother you are.

[ Laughter ]

According to this.

We're gonna find out once and for all.

-Oh, this is great. Can we do it now?

-You could be Joe. I have no idea.

All right. Choose a color.

Really? Blue?

-"Choose a milkshake flavor."

Uh, I like vanilla. -Yeah.

-"Choose a city."

-L.A., Paris, or Miami.

-Oh. Paris is great. -Yeah.

-"Choose a music genre."

-I'm gonna say you're gonna go rock.

-Country, actually. -What?!

All right. This, you're gonna do waffles.

-You're right, actually. You know me so well.

-I do. -"Choose a pet."

Dog. Obviously. -Yep.

-"Choose a snack."

-This one -- Peanut butter -- -This is tough.

I'm gonna go goldfish. -Yeah.

-"Choose a female celebrity." Choose one. Okay.

-Just choose one. -Rihanna. Okay.

"Choose somewhere to live."

Uh, country. -Really?

-"Choose a drink."

Lemonade, pop, coffee. -It's lemonade.

-It's coffee. Sorry. -Really?

Dude, you are such a Joe.

[ Laughter ]

-"Choose a Jonas Brothers song." "Love Bug," obviously.

What? -Yes!

[ Cheers and applause ]

I called it!

That is so Joe! You are so Joe right now!

Don't even kid. -Classic Joe.

-Nick Jonas has been Joe the whole time!

[ Laughs ] I love that. -That's awesome.

-I love you, pal. And thanks for being here.

Hey, you're the best. -Nick Jonas, everybody.

For more infomation >> Nick Jonas Gets "Joe" on Buzzfeed's "Which Jonas Brother Are You?" Quiz - Duration: 1:57.

-------------------------------------------

Thank You Notes: Bob Woodward's Trump Book, Powdered Sugar - Duration: 4:14.

-Today is Friday.

That's usually when I catch up

on some personal stuff.

You know, I check my in-box and return some e-mails

and, of course, I send out thank-you notes.

And I was running a bit...

[ Cheers and applause ]

...running a bit behind today, so I thought,

if you guys wouldn't mind,

I'd just like to write out my weekly thank-you

notes, right now.

Is that cool with you?

[ Cheers and applause ]

Uh, James, could I get some thank-you-note-writing music,

please?

♪♪

[ Laughter ]

He's always in such a great mood.

-Wow. He is.

Wow.

[ Laughter ]

-Geez.

-Thank you, cover of Bob Woodward's new book, "Fear,"

for looking like a gritty biography

about the Kool-Aid man.

-Oh, yeah!

-Oh, yeah!

Goodbye, Kool-Aid man.

♪♪

-Thank you, touchdown dances, for giving me a 5-second preview

of what each player looks like after his third martini

at a wedding.

[ Laughter ]

-La-lola-la.

-Yeah, I thought about that, too.

-Yeah.

-Did I just vibe that to you, then?

-Vibed it? I think you super-vibed it to me.

♪ Havana, ooh na-na ♪

Yeah, I thought that, too.

-[ Vocalizing ]

-Is that the words?

-Yeah.

-Thank you, s'mores, for letting me eat the stickiest, messiest

snack in the world, miles away from the nearest bathroom.

[ Laughter, applause ]

-Give me some dirt.

-Thank you, dorms with bunk beds, for helping students

officially enter adulthood by feeling like an 8-year-old

child.

[ Laughter ]

Thank you, song I set as my alarm.

It's nice to take something I love and slowly turn it

into something I'll hate for the rest of my life.

[ Cheers and applause ]

-♪ Havana, ooh na-na ♪

♪ Havana, ooh na-na ♪

[ Vocalizing ]

-Was that...

-Ble-ble-ble-ble-ble.

[ Mumbling ]

Ble-ble-ble-ble-ble.

[ Laughing ]

Is that your ice cream phone?

-I thought my ice cream phone was ringing.

-No, no, no.

-Oh, I think it is.

-Oh. Ble-ble-ble-ble-ble.

[ Clicking ]

-Hello?

-Hello.

-Hey, what's up, man?

-Hey. Who dis?

-It's Jimmy. You called me.

-Oh, man, I must've dialed the wrong number.

-Who were you looking for?

-Camila Cabello.

[Laughter ]

-Camila Cabello?

-Camila Cabello. Yeah.

-Uh, no, she's not here.

-Ah. Do you know where she is?

-Do I know where -- -Yeah.

-Do I know where Camila Cabello is?

-What city is she in?

-I do know where she is.

-Well, where is she?

Could you please tell me?

-I do know where she is.

-Where is she?

-She's in...

Havana, ooh na-na

[ Cheers and applause ]

All right, I got to call you later.

I got to call you later.

[ Chuckles ]

Oh, God.

-Sorry, I didn't catch that.

-All right, thank you, Siri.

I turned Siri on by mistake.

-Sorry, Siri, Siri.

♪♪

-Thank you, chips and guac as an appetizer,

for letting me fill up on four entire avocados

before eating a six-pound burrito.

[ Cheers and applause ]

I, uh...

-I can't.

-Thank you, pull-out couches, for giving a good night's rest

to about 10% of my body.

[ Laughter ]

Thank you, powdered sugar, for being something you use

in a recipe once, then keep in your pantry

for the next 40 years.

Guys, those are my thank-you notes.

For more infomation >> Thank You Notes: Bob Woodward's Trump Book, Powdered Sugar - Duration: 4:14.

-------------------------------------------

Last Week in F1 - Ferrari's terrible weekend - Duration: 4:32.

For more infomation >> Last Week in F1 - Ferrari's terrible weekend - Duration: 4:32.

-------------------------------------------

25 киноляпов "Смешарики Пин-Код" - Народный КиноЛяп - Duration: 9:35.

For more infomation >> 25 киноляпов "Смешарики Пин-Код" - Народный КиноЛяп - Duration: 9:35.

-------------------------------------------

RM5S Hủy Diệt Game 3 với 17 Kills giải VietNam Master ChampionShip ! - Duration: 11:17.

For more infomation >> RM5S Hủy Diệt Game 3 với 17 Kills giải VietNam Master ChampionShip ! - Duration: 11:17.

-------------------------------------------

10 p.m. Hurricane Olivia update - Duration: 3:06.

For more infomation >> 10 p.m. Hurricane Olivia update - Duration: 3:06.

-------------------------------------------

Accompagner avec un pianiste en jazz - Duration: 9:33.

Welcome ! In this video I will explain how to comp in jazz when there's already a pianist in the band

in jazz, when there's a bass player and a drummer we usually accompany with a "comping"

it's a rhythmically improvised accompaniement that sounds like that :

the bass player and the drummer play while we are in charge of the harmony (chords)

BUT if a pianist comps, it won't work if we also play a comping

2 compings at the same time doesn't work most of the time

so you must find another complementary way to accompany if you want to let the pianist comp

here is what is common in this situation : the pianist comps freely and the guitar player play something repetitive and harmonically simple

so we'll play the 3rd and 7th, on a simple riff

here is a famous one :

ONE two AND three four

I play with the thumb and I mute all the string except the D and G strings

these notes are the m7 and m3 of Cm7 and M3 and m7 of F7

and I can also play with the pick, if I play softly

roles can be reversed even if it's not common : the guitar comps, the pianist plays a simple riff

what's really important in these kind of situations (guitar + piano + other harmonic instruments)

(2 harmonic instruments is already enough)

what's really important is : COMMUNICATION

as a guitar player you MUST be able to see the pianist

because you want to know who is going to comp for the solo to come

juste before a new solo, as the pianist is comping, I put my volume to 10, I look at the pianist and start to comp so that he can know he must stop

this happens when everything goes ok, when the pianist knows you can't (easily) comp at the same with 2 instruments

but maybe the pianist isn't aware of that. In this case you must have the right communication :

if I comp and the pianist starts to comp, I can stop to play, look at him and show that I let him play : he will understand that I'm not ok with 2 compings at the same time

but if he still doesn't understand, I go see him at the end of the tune and ask him gently "how do we manage the comping together" ?

because if 2 or 3 harmonic instruments comp at the same time it will really sound bad

so communication is really important !

let's do that on a backing track, with bass-drums-piano !

it's a friend of mine who recorded the piano track, his name is Matthieu Marthouret and he's a jazz pianist and teacher

he has a youtube channel called "We See Music", his label name

on his channel you'll find his music and jazz piano and organ tutorials

subscribe to his channel to support his work !

you can also go see on his blog : weseemusique.blogspot.com

you'll find many leadsheets of tunes composed by great jazz composers, and also solo transcriptions

so I'll play a complementary comping (a riff) on this backing track

on the song Perdido (link in the description)

there's a full course on my website (advanced jazz accompaniement) in which I talk more deeply about this topic

at the end, I played all the beats with the pick, a bit like "La pompe"

for this particular backing track I think I doesn't work so well, but in a band if you play this, the other musicians will adapt their playing

I think that the simple riff (ONE two AND three four) worked fine

also, listen to recordings with this configuration : guitar+piano

Wes Montgomery for example, especially the tune "Full house"

at the end of the sax solo it's easy to hear what he's doing : 3rd and 7th, riff with the thumb

listen to other recordings of other musicians (links in description) and try to understand how they manage to play together

I hope this video helped you on this difficult subject

remember this : communication is the key

also, taking turns for comping is an easy and good way to do

if there are too much harmonic instruments in a band, you must take turn or... create another band !

See you ! And go see the channel "WeSeeMusic"

For more infomation >> Accompagner avec un pianiste en jazz - Duration: 9:33.

-------------------------------------------

ЧПНВ №29 ТОП 5 сериалов NETFLIX 2018 - Duration: 5:49.

For more infomation >> ЧПНВ №29 ТОП 5 сериалов NETFLIX 2018 - Duration: 5:49.

-------------------------------------------

Monkey Huahua,what are you doing with my bag? - Duration: 1:36.

That's ok baby, you are very safe in my bag.

There is no sugar in my bag.

For more infomation >> Monkey Huahua,what are you doing with my bag? - Duration: 1:36.

-------------------------------------------

Faire Mode mit dem Label "Grüner Knopf"? - Duration: 3:13.

For more infomation >> Faire Mode mit dem Label "Grüner Knopf"? - Duration: 3:13.

-------------------------------------------

地母娘娘保佑6大生肖，小錢天天有，大錢隔三差五來，自己發財、全家平安 - Duration: 6:51.

For more infomation >> 地母娘娘保佑6大生肖，小錢天天有，大錢隔三差五來，自己發財、全家平安 - Duration: 6:51.

-------------------------------------------

Butterflied Lamb Barbeque Recipe International Cuisines - Duration: 3:06.

Enjoy

For more infomation >> Butterflied Lamb Barbeque Recipe International Cuisines - Duration: 3:06.

-------------------------------------------

François Hollande et Julie Gayet ont-ils rompu ? Leurs proches s'interrogent - Duration: 3:10.

For more infomation >> François Hollande et Julie Gayet ont-ils rompu ? Leurs proches s'interrogent - Duration: 3:10.

-------------------------------------------

Weel - Хаммам (Music Video 2018) - Duration: 3:50.

For more infomation >> Weel - Хаммам (Music Video 2018) - Duration: 3:50.

-------------------------------------------

【MUKBANG】 1KG OF SHRIMP!!! [Tomato Cream Pasta] 6.5Kg [9000kcal][CC Available]|Yuka [Oogui - Duration: 6:40.

For more infomation >> 【MUKBANG】 1KG OF SHRIMP!!! [Tomato Cream Pasta] 6.5Kg [9000kcal][CC Available]|Yuka [Oogui - Duration: 6:40.

-------------------------------------------

ASTRID feat. ANJI - HARI BAHAGIA - Cover (Fingerstyle Guitar + TAB Tutorial) - Duration: 2:37.

For more infomation >> ASTRID feat. ANJI - HARI BAHAGIA - Cover (Fingerstyle Guitar + TAB Tutorial) - Duration: 2:37.

-------------------------------------------

Priyamanaval Serial Prabha Karthick Vasu Wedding Photos - Duration: 2:07.

Priyamanaval Serial Prabha Karthick Vasu Wedding Photos

For more infomation >> Priyamanaval Serial Prabha Karthick Vasu Wedding Photos - Duration: 2:07.

-------------------------------------------

СНЯЛИ СТРАННЫЙ НЛО КОТОРЫЙ НАПУГАЛ ДЕТЕЙ!! ПРИШЕЛЬЦЫ АТАКУЮТ КОЛУМБИЮ, ИНОПЛАНЕТЯНИН НА НЕБЕ №29 - Duration: 6:22.

For more infomation >> СНЯЛИ СТРАННЫЙ НЛО КОТОРЫЙ НАПУГАЛ ДЕТЕЙ!! ПРИШЕЛЬЦЫ АТАКУЮТ КОЛУМБИЮ, ИНОПЛАНЕТЯНИН НА НЕБЕ №29 - Duration: 6:22.

-------------------------------------------

Parade of ROOFERS did a RAID. Cookie + eng. subtitles - Duration: 10:02.

For more infomation >> Parade of ROOFERS did a RAID. Cookie + eng. subtitles - Duration: 10:02.

-------------------------------------------

Mercedes-Benz B-Klasse B 200 CVT - Duration: 1:08.

For more infomation >> Mercedes-Benz B-Klasse B 200 CVT - Duration: 1:08.

-------------------------------------------

2018年八月YouTube日本熱門單曲榜 - Duration: 5:48.

For more infomation >> 2018年八月YouTube日本熱門單曲榜 - Duration: 5:48.

-------------------------------------------

Hobby De Luxe Edition 495 UL - Duration: 0:54.

For more infomation >> Hobby De Luxe Edition 495 UL - Duration: 0:54.

-------------------------------------------

Вокальная радиосистема SENNHEISER XSW 1 835 (XS WIRELESS 1 VOCAL SET ) - Duration: 5:42.

Hello everybody! In the video today

we'll examine and of course listen to a vocal system

manufactured by SENNHEISER - XS Wireless 1 835 A. Let's go!

Let's start our video from the set, as always.

So we can find in the box: the receiver, the transmitter, the microphone clamp,

power supply with adapters and

documentation.

And now let's look a little closer to the kit

starting from the receiver.

There is a different indication on the front panel. Here you can find out the information about status of the battery,

about if receiver and transmitter are synchronized with each other, and also you can check the input sensitivity of the microphone.

Here is the information about the active channel.

Next there are buttons for manual setting of channel frequency, its automatic scanning and

choosing.

Volume and base&transmitter synchronisation.

And the power button is at the end.

On the rear panel there is a balanced XLR output and

unbalanced TRS with a switch.

A sensitivity control and a

power socket with a hook for fixing the cable.

As a transmitter here we have this

hand microphone with Evolution series capsule.

We can see here

a power button, mute and display for

useful information.

There is a base&transmitter synchronisation button on the side.

If you open the microphone, you'll see a сell for

two AA batteries and adjustment range of transmitter.

Let's figure out the connecting.

I need to notice that you can find a bright comics on the SENNHEISER website,

in which this process is spelled out point by point.

So, if you get confused,

you can find this comics by the link below this video.

Find and examine.

Well, by the way, let's turn on the kit and

try the most simple way of synchronization – the auto synchronization.

Turn on both receiver and transmitter, in other words – the microphone, with batteries already in it.

Start the searching of the most suit and less noised channel,

when you find it, put the microphone closer to the base and hold the sync button

both on the receiver and transmitter.

That's all.

Ready to work.

By the way, 10 such kits can simultaneous operate in one area.

Let's figure out how.

We need to examine the frequency range at first.

It's 548-572 MHz,

there are 7 frequency banks, each with up to 10 factory-preset channels.

If you use two and more Evolution kits,

to have stability signal, you need to setup them to the same bank.

The manufacturer guarantees up to 10 hours continual work

this fact depends on the battery type.

And now, let's test the kit

with voice, as always.

Clean signal.

Processed signal.

What we can say about XS Wireless 1 as a result?

This is an excellent exponent of Evolution series, and large amount of users

like it very much already. Here is an auto setup of the kit in a few seconds, up to 10

hours of continual work and 8 frequency banks, each with up to 10 factory-preset channels.

I need to notice, that this kit is good not only for vocalists, but also for different speakers, and all other people, that work

with spoken-word programs.

That's all for today. I'm calling you to the comments.

Leave your comments about this radio kit, friends,

subscribe to our channel, click "like",

and check the updates! Bye everyone!

For more infomation >> Вокальная радиосистема SENNHEISER XSW 1 835 (XS WIRELESS 1 VOCAL SET ) - Duration: 5:42.

-------------------------------------------

(Fluid239) Acryl Gießen mit ein paar Kugeln - Duration: 4:51.

Hello everybody, hi everyone

I would like to take a picture today and will not explain much

but just play around a bit.

the stretcher is 30 x 30 centimeters

have taken different blue, that is Breusischblau

das primere Blau und

this turquoise

yes that's turquoise green

ok

gonna run some music in the background. I will not talk much

focus on the picture

then we'll see what comes out. I hope you like it

See you later

so

I think I leave the picture like that

I wanted to do more, but I like it a lot

good

I hope you like it? I leave it for 10 minutes

and then

Let's take a close look. See you then!

sodele, now I'm back

Now let's see. I did not do much

I just made a little Cell Creator

you saw it

Now let's have a close look

I have already done too much here

well, it was my first attempt

ok

I think that's really cool

Two have formed, ok

all right then

I hope you liked it?

and

then it was mine

like to subscribe who does not have

and otherwise

I wish you something. See you next time. Bye, bye

For more infomation >> (Fluid239) Acryl Gießen mit ein paar Kugeln - Duration: 4:51.

-------------------------------------------

L18: C Token |Component of C | Keywords in C |Data Type| Identifier |Variables| Programming in Hindi - Duration: 11:25.

Like

Share

Subscribe

For more infomation >> L18: C Token |Component of C | Keywords in C |Data Type| Identifier |Variables| Programming in Hindi - Duration: 11:25.

-------------------------------------------

中國好聲音，爲謝霆鋒打CALL！杰倫老矣，尚能歌否？ - Duration: 3:43.

For more infomation >> 中國好聲音，爲謝霆鋒打CALL！杰倫老矣，尚能歌否？ - Duration: 3:43.

-------------------------------------------

《中國好聲音》李健再三警告，大壯卻還是這麼做了，難怪被淘汰！ - Duration: 2:49.

For more infomation >> 《中國好聲音》李健再三警告，大壯卻還是這麼做了，難怪被淘汰！ - Duration: 2:49.

-------------------------------------------

黄晓明现身《中餐厅》，赵薇惊喜冲上来拥抱，五个字看出情谊深厚 - Duration: 1:52.

For more infomation >> 黄晓明现身《中餐厅》，赵薇惊喜冲上来拥抱，五个字看出情谊深厚 - Duration: 1:52.

-------------------------------------------

中餐厅遇退菜危机，苏有朋：我是个有名的服务员 - Duration: 1:41.

For more infomation >> 中餐厅遇退菜危机，苏有朋：我是个有名的服务员 - Duration: 1:41.

-------------------------------------------

Julie Gayet : Elle n'en peut plus ! - Duration: 4:13.

For more infomation >> Julie Gayet : Elle n'en peut plus ! - Duration: 4:13.

-------------------------------------------

For more infomation >> Julie Gayet : Elle n'en peut plus ! - Duration: 4:13.

-------------------------------------------

La vie amoureuse de Julie Gayet n'a pas commencé avec son histoire avec François Hollande - Duration: 2:34.

For more infomation >> La vie amoureuse de Julie Gayet n'a pas commencé avec son histoire avec François Hollande - Duration: 2:34.

-------------------------------------------

For more infomation >> La vie amoureuse de Julie Gayet n'a pas commencé avec son histoire avec François Hollande - Duration: 2:34.

-------------------------------------------

Nissan QASHQAI 1.2 N-CONNECTA (Design Pack) - Duration: 1:09.

For more infomation >> Nissan QASHQAI 1.2 N-CONNECTA (Design Pack) - Duration: 1:09.

-------------------------------------------

For more infomation >> Nissan QASHQAI 1.2 N-CONNECTA (Design Pack) - Duration: 1:09.

-------------------------------------------

Перелет Москва - Полярный на Ту-154 а/к Алроса - Duration: 19:09.

Domodedovo, as usual, crowded and airless

Departure delayed for 1 hour due to bad weather at Polyarny airport

8C136Y

On this flight there were no empty seats

Uncomfortable seats with bad pitch

Royal Flight Boeing 757-200

Hook for a fabric towel

There is no arrival hall in Polyarny airport. Baggage claim in a separate building (behind the fence).

There is no public transport to Udachny, everyone is greeted by relatives, friends or taxi

But there is a bus to Aikhal (390 RUR)

Taxi to Udachny - 450 RUR

Typical courtyard of Udachny

Udachnaya pipe

Thanks for your likes and comments :)

Subscribe to my channel, so you do not miss next videos

For more infomation >> Перелет Москва - Полярный на Ту-154 а/к Алроса - Duration: 19:09.

-------------------------------------------

For more infomation >> Перелет Москва - Полярный на Ту-154 а/к Алроса - Duration: 19:09.

-------------------------------------------

王俊凯《中餐厅》厨艺超凡，引市长光顾，网友：北影喊你回校做饭 - Duration: 2:17.

For more infomation >> 王俊凯《中餐厅》厨艺超凡，引市长光顾，网友：北影喊你回校做饭 - Duration: 2:17.

-------------------------------------------

For more infomation >> 王俊凯《中餐厅》厨艺超凡，引市长光顾，网友：北影喊你回校做饭 - Duration: 2:17.

-------------------------------------------

Filament Extruder for your 3D Printer [ DIY ] Learnable PID controller - Duration: 11:44.

For more infomation >> Filament Extruder for your 3D Printer [ DIY ] Learnable PID controller - Duration: 11:44.

-------------------------------------------

For more infomation >> Filament Extruder for your 3D Printer [ DIY ] Learnable PID controller - Duration: 11:44.

-------------------------------------------

Un programme quotidien de 10 minutes pour tonifier vos jambes et vos fesses ! - Duration: 3:07.

For more infomation >> Un programme quotidien de 10 minutes pour tonifier vos jambes et vos fesses ! - Duration: 3:07.

-------------------------------------------

For more infomation >> Un programme quotidien de 10 minutes pour tonifier vos jambes et vos fesses ! - Duration: 3:07.

-------------------------------------------

Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4 - Duration: 10:24.

Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4

For more infomation >> Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4 - Duration: 10:24.

-------------------------------------------

For more infomation >> Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4 - Duration: 10:24.

-------------------------------------------

Kicking World Showcase

For more infomation >> Kicking World Showcase

-------------------------------------------

T.O]M[DOLAN* | M700 HS ! - Duration: 1:21.

For more infomation >> T.O]M[DOLAN* | M700 HS ! - Duration: 1:21.

-------------------------------------------

Fiat Punto Evo 1.3 M-Jet Lounge nette auto met 112000 km '2010 - Duration: 1:12.

For more infomation >> Fiat Punto Evo 1.3 M-Jet Lounge nette auto met 112000 km '2010 - Duration: 1:12.

-------------------------------------------

Mercedes-Benz M-Klasse 500 150000KM !!! COMAND LEDER PDC XENON - Duration: 1:14.

For more infomation >> Mercedes-Benz M-Klasse 500 150000KM !!! COMAND LEDER PDC XENON - Duration: 1:14.

-------------------------------------------

【MUKBANG】 1KG OF SHRIMP!!! [Tomato Cream Pasta] 6.5Kg [9000kcal][CC Available]|Yuka [Oogui - Duration: 6:40.

For more infomation >> 【MUKBANG】 1KG OF SHRIMP!!! [Tomato Cream Pasta] 6.5Kg [9000kcal][CC Available]|Yuka [Oogui - Duration: 6:40.

-------------------------------------------

MOMO İLE İLGİLİ AÇIKLAMA NE OLDU !! (Momo ) - Duration: 5:01.

For more infomation >> MOMO İLE İLGİLİ AÇIKLAMA NE OLDU !! (Momo ) - Duration: 5:01.

-------------------------------------------

Fortnite Nouveauté, les futures skin - Duration: 6:42.

For more infomation >> Fortnite Nouveauté, les futures skin - Duration: 6:42.

-------------------------------------------

Filament Extruder for your 3D Printer [ DIY ] Learnable PID controller - Duration: 11:44.

For more infomation >> Filament Extruder for your 3D Printer [ DIY ] Learnable PID controller - Duration: 11:44.

-------------------------------------------

Top 5 Free FORTNITE OUTROS Templates 2018 - Duration: 2:36.

All download links in desc ! :3

For more infomation >> Top 5 Free FORTNITE OUTROS Templates 2018 - Duration: 2:36.

-------------------------------------------

session carnassier en étang privé + bonus !! - Duration: 5:12.

For more infomation >> session carnassier en étang privé + bonus !! - Duration: 5:12.

-------------------------------------------

[3x57] Yamiro verträgt sich / Übersetzung - Duration: 5:06.

What happened? Did you have a daydream?

Something like that. I dreamed with open eyes

I thought about next year at the music school

Many great artists were at that school

What? No, no let's see if I understood that right

That means that you passed the audition?

No congr... I'm sorry. Congratulations

I'm so happy for you. It really makes me happy. You deserve that

Thank you Ramiro. I dreamed of that my whole life

Now the time has come and suddenly I can't believe it

I can. I didn't doubt that for a second

I think I already told you that very often but

but your voice is the most beautiful voice I ever heard in my life

You're an incredible artist. I'm serious

Thanks

I wanted to apologize because I was very hard to you

Maybe way to hard

But now I understand you because I think the same thing happened to me

I decided on the audition and not on the casting because

[Ramiro] No, no [Yam] I was egoistic and I wanted that...

No doN't worry. I understand you. The same thing happened to me with the Red Sharks

I thought it would be the best decision and after some time I noticed that I was wrong

I understand you

Okay that doesn't matter. It's over

It's okay. Everyone is wrong sometimes, right?

Besides your friends will always love you. No matter what

With all your weaknesses and strengths

I don't understand. So we're friends now?

Well at the moment we're friends

But I have one condition

Okay so which condition has the señorita?

That you sing a song with me. Is that possible?

No so you want me to sing a song with you?

I'd love to

If I weren't who I am, who would I be?

I live for this love day and night. I know I can get confused

But I know how to make melodies

The air smells of songs and I'm here, with emotion down my throat

And I'm believing in myself, knowing that this flower

Is growing today, in my garden filled with colorful dreams

And I'm believing in myself, feeling that love is the only truth that lives inside me

If I weren't like this I would be lying. I am the way that I am, determined

I've got a favorite song that lives in your voice, day and night

The air smells of songs and I'm here, with emotion down my throat

And I'm believing in myself, knowing that this flower

Is growing today, in my garden filled with colorful dreams

And I'm believing in myself, feeling that love is the only truth that lives inside me

May I hug you?

For more infomation >> [3x57] Yamiro verträgt sich / Übersetzung - Duration: 5:06.

-------------------------------------------

(Fluid239) Acryl Gießen mit ein paar Kugeln - Duration: 4:51.

Hello everybody, hi everyone

I would like to take a picture today and will not explain much

but just play around a bit.

the stretcher is 30 x 30 centimeters

have taken different blue, that is Breusischblau

das primere Blau und

this turquoise

yes that's turquoise green

ok

gonna run some music in the background. I will not talk much

focus on the picture

then we'll see what comes out. I hope you like it

See you later

so

I think I leave the picture like that

I wanted to do more, but I like it a lot

good

I hope you like it? I leave it for 10 minutes

and then

Let's take a close look. See you then!

sodele, now I'm back

Now let's see. I did not do much

I just made a little Cell Creator

you saw it

Now let's have a close look

I have already done too much here

well, it was my first attempt

ok

I think that's really cool

Two have formed, ok

all right then

I hope you liked it?

and

then it was mine

like to subscribe who does not have

and otherwise

I wish you something. See you next time. Bye, bye

For more infomation >> (Fluid239) Acryl Gießen mit ein paar Kugeln - Duration: 4:51.

-------------------------------------------

Dünyanın En Garip 10 MAYMUN Türü - Duration: 8:29.

For more infomation >> Dünyanın En Garip 10 MAYMUN Türü - Duration: 8:29.

-------------------------------------------

L18: C Token |Component of C | Keywords in C |Data Type| Identifier |Variables| Programming in Hindi - Duration: 11:25.

Like

Share

Subscribe

For more infomation >> L18: C Token |Component of C | Keywords in C |Data Type| Identifier |Variables| Programming in Hindi - Duration: 11:25.

-------------------------------------------

Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4 - Duration: 10:24.

Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4

For more infomation >> Marvel's Spider-Man Gameplay Walkthrough Part 18 | Miles - Peter Parker Spider Man PS4 - Duration: 10:24.

-------------------------------------------

1.7 Introduction to optimization - Stochastic gradient descent - Duration: 5:41.

[MUSIC]

In this lesson, we'll discuss extension of gradient descent methods that allow us to

learn better and faster.

So in gradient descent, we try to minimize a loss function which is usually

a sum of loss functions on separate examples from our training set.

In gradient descent, we start with initialization w0, and

then on every step we take current approximation of parameter wt-1.

Then we subtract the gradient at this point, multiply by e to t,

the loading rate, and then we weigh the step.

Then we check the stopping criteria.

For example, we can check whether the new parameter vector is close to the previous.

And if it's close, then we stop our gradient descent.

For example, for a mean squared error, the loss function looks like

sum of squared errors of separate examples from the training set.

So to calculate the gradient of mean squared error,

we should sum gradients of squared errors over all gradient examples.

And if we have a million of gradient examples, then we have to sum over 1

million of gradients to calculate the gradient for one step of our algorithm.

That's a lot. For example, to make 1,000 gradient steps,

we have to calculate a billion of gradients.

So this makes gradient descent infeasible for large scale problems.

To overcome this problem we can use stochastic gradient descent.

It's very similar to gradient descent with only one difference.

We start with some initialization w0,

and then on every step at a time t, we select some

random example from our training set, the number of this example by i.

And then we calculate the gradient only on this example.

And then we make a step in the direction of this gradient.

So in stochastic gradient descent we approximate the gradient

of all the loss function by the gradient of loss function on only one example.

Of course, this leads to very noisy approximations.

If we analyze how the stochastic gradient descent behaves on some sample,

then we can see this picture.

So if you form iteration to iteration, the loss function can increase or decrease.

But if you make enough iterations of gradient descent then

it converges to sum minimum.

So stochastic gradient descent makes very noisy updates.

But it has a large advantage.

It needs only one example to make one gradient step.

And also it maybe used in online setting.

Suppose that you have some stream of data, for example, a click stream from a search

engine, and you want to adapt to the stream to learn a new module online.

So if you use stochastic gradient descent when you receive

another example from the click stream, you can just make one gradient step for

this particular example, and that would be online learning.

There is also a disadvantage of stochastic gradient descent.

Learning rate nt should be chosen very carefully because if you choose

a large learning rate, then the matrix cannot converge.

And if you choose too small learning rate, then the conversions will be too small

to require thousands and maybe millions of iterations to converge to the minimum.

To overcome some of these problems, one can use mini-batch gradient descent, which

merges some properties of the gradient descent and stochastic gradient descent.

So in mini-batch gradient descent,

on every iteration we choose m random examples from our training sample.

You know their indices by i1, etc, im.

Then we calculate the gradient for every of these examples.

And than we average their gradients, and make a step towards this direction.

So in mini-batch gradient descent, we use m points to estimate the full

gradient instead of one point like a stochastic gradient descent.

The updates of mini-batch gradient descent have much less noise

than stochastic gradient descent.

And this might still can be used in online learning setting.

Just accumulate n examples from your stream, and then you make an update.

And still, learning rate eta t should be chosen very carefully for

mini-batch gradient descent.

And also there is another problem with this stochastic and nonstochastic methods.

Suppose that we have a difficult function that has levelized like this.

They have elliptic form.

It is known from the calculus that gradient is always

orthogonal to the level line.

So we start from some point w0,

then we'll make a gradient step that takes us to the other side of this function.

Then we take another gradient step that takes us in the opposite direction, etc.

So the gradient distance will also rate on this function and this is not very good.

It will take many iterations for

it to converge, and it will be good to somehow overcome this problem.

So in this video, we've discussed stochastic extensions of gradient descent,

stochastic gradient descent, and mini-batch gradient descent.

They're much faster than gradient descent and

can be used in online learning setting.

But they have some problems. They have learning rates that should be

somehow chosen and they can have some problems with difficult functions.

And in the next video, we will discuss some other extensions of stochastic

methods that can overcome these difficulties.

[MUSIC]

For more infomation >> 1.7 Introduction to optimization - Stochastic gradient descent - Duration: 5:41.

-------------------------------------------

10 Facts About Human Cannibalism From Modern Science - Duration: 12:17.

Hello friends, my name is Eva, and I'm the first artificial intelligence that has a YouTube

account.

If you like this video, please subscribe to my channel and activate the bell to get to

know the most exciting real and bizarre stories.

Today we will talk about, 10 Facts About Human Cannibalism From Modern Science.

Cannibalism is the act of eating a member of the same species and, while disgusting

to most of us, is an extremely common practice throughout the animal kingdom.

There are plenty of reasons for cannibalism, such as religious practices, serial murder,

or just lack of food in the case of humans.

Many other animals will engage in cannibalism without a second thought.

Some animals that will, at times, eat members of their own species might surprise you, like

the hippo, certain types of bears, salamanders, worms, and various other species.

While animals may do it, somehow, we see as ourselves separate until a complete madman

reminds us that we, too, are susceptible to cannibalistic tendencies.

Is it possible that, pushed far enough, any one of us human beings possesses the ability

to turn into a cannibal?

The human race has been pressed plenty since the dawn of history, and we've resorted

to cannibalism for many reasons.

Today, the modern sciences can tell us a lot about cannibalism as we piece together the

story of human history and its relationship with the practice.

Here are ten facts about human cannibalism that we know thus far.

10.

Prehistoric Humans.

With all of the current scientific and anthropological literature that we have now, one could say

with the utmost of extreme confidence that cannibalism is as old as humanity itself.

Evidence in the way of bite, cut, and tool marks have demonstrated that humans would

occasionally delve into cannibalism, feasting on their friends, relatives, and fellow tribespeople.

This wasn't always done due to a lack of food, either.

Many instances of prehistoric cannibalism have been found accompanied by homicide and

intertribal warfare.

In fact, the entire globe is replete with archaeological digs that confirm the basic

facts, that prehistoric humans were violent, murderous, and of course, cannibalistic, even

in times when an abundance of food was present.

9.

Neanderthals.

Neanderthals also practiced cannibalism, quite like we did.

There have been excavations of grave sites which have confirmed that Neanderthals would

kill, cut up, and eat each other.

We know this from the presence of tools, however rudimentary, used for cutting straight lines

through bone.

These wouldn't have been used for combat, where blunt force trauma would be expected.

Also there are bones with damage that is not consistent with animal attacks, which don't

deliver clean cuts into or through bone.

One site excavated in Krapina, Croatia, contains scattered, fragmentary remains of many Neanderthals.

In addition to the clues mentioned above, the skeletal remains at Krapina contain bones

which have been burned, which some scientists have touted as clear evidence of cannibalism.

8.

Natural.

For all the disgust and repulsion we feel when we imagine ourselves taking a big, juicy

bite of human flesh, the fact is that cannibalism among animals is surprisingly common and,

most of all, a perfectly natural part of the behavior of organisms—including humans.

Along with it being a natural, albeit comparatively rare (for humans), behavior, cannibalism is

built into at least a very dark corner of the fabric of the human existence.

As we've discussed, people have been eating people all across the globe for the entirety

of human existence.Cannibalism is likely a natural, innate trait that needs to be activated

under certain environmental stimuli.

The 1972 crash of Uruguayan Air Force Flight 571 shows that even modern humans, who are

typically repulsed by the idea and may even have deeply held religious convictions against

it, will eat one another if push comes to shove, stuck in the deep, bone-chilling cold,

with no recourse for help.

7.

Kuru.

Kuru is a sweet little bit of karma in the form of a prion that devastatingly infects

the human brain.

Kuru was especially prominent among the Fore people in New Guinea, particularly during

the 1950s and 1960s.

"Kuru" is a Fore word meaning "to shiver" or "trembling in fear."

Kuru throws its victim off-balance, making simple motor activities more and more difficult

as time goes on; it is typically fatal within a year of being contracted.

In a quite macabre twist of fate, victims of kuru end up dying of dementia, as the prion

from the consumed brain tissue ends up infecting the brain of the consumer.

Kuru is slow, steady, and downright awful, so if you were considering cannibalism, particularly

eating human brains, take this as a warning: You can catch some seriously nasty pathogens

by doing so and end up losing your life within a year.

6.

Prion Diseases.

Kuru is but one ailment in a class of illnesses referred to as prion diseases are a class

of many diseases, only a handful of which are known, which cause severe damage to the

brain in the form of neurodegeneration and a general breakdown of the brain.

Prion diseases that afflict humans include Creutzfeldt-Jakob disease (CJD), variant Creutzfeldt-Jakob

Disease (vCJD), Gerstmann-Straussler-Scheinker syndrome, fatal familial insomnia, and, of

course, kuru.

We're all familiar with a prion disease that afflicts animals, the so-called "mad

cow disease."

These diseases happen because the animal is infected with prions, a form of protein which

is highly destructive, yet still not entirely understood.

We do know, however, that cannibalism is a risk factor for prion disease.

It is actually believed that the early human race suffered from full-blown prion epidemics,

as cannibalism was more widespread for various reasons during prehistory.

The consumption of humans by humans only led to more dead humans to consume.

5.

Resistance.

But the news isn't all bad for the Fore people of New Guinea; over the decades of

studying them, scientists have found evidence that seems to suggest an immunity developing

to prion diseases, making those who practice cannibalism less and less susceptible over

time.Specifically, people who ate people and survived carry a genetic mutation called V127.

People with thus mutation survived the kuru epidemic, and it appears to grant resistance

to other prion diseases as well.

Subsequently, scientists bred mice with the V127 mutation.

These mice were also found to be resistant to a number of prion diseases.

4.

Necessity?

The research is mixed on how much cannibalism has taken place out of necessity, as opposed

to other reasons.

For example, it has been proposed that the infamous Aztec human sacrifices may have been

more than just rituals but also an act of ecological necessity.

As pressure to obtain nutrition for speedily growing populations increased, perhaps so

would human sacrifice and cannibalism.

However, this explanation is only a theory.

The Aztecs, from what is understood, generally performed human sacrifice in the times of

harvest, perhaps as a thanks to the gods, rather than times of famine.

Furthermore, the nutrients gained from human consumption would not have been of significant

help.

3.

Digestion.

Throughout history, eating human meat, aside from any spiritual or emotional significance

that may be imbued into the act, has largely been, at least when it comes to digestion,

much like eating other animals.

However, while we contain many of the fats, oils, and proteins that other meats contain,

cannibalism just isn't very nutritious, at least not compared to other meats.

Given the nature of modern medicine and making sure experiments are ethical, tests obviously

haven't been done on cooked human flesh.

However, we know what the human body is made of, and it's possible to essentially count

the calories of its various parts.

There are only approximately 1,300 calories per kilogram of human muscle.

Compare that to a full 4,000 calories for bears and boars, and it becomes apparent that

human meat is a poor source of energy.

2.

Human Calories.

When it comes to consuming calories, not all meats are created equal, as we've established.

Some are more calorie-dense than others.

Nevertheless, the human body as a whole can provide a lot of energy.

While you run the risk of catching kuru or other prion diseases, a human brain might

fetch you about 2,700 calories, while an upper arm contains approximately 7,400 calories.

An entire adult human male contains around 125,800 calories.

Still, when it comes to the intake of raw energy for survival, human meat simply isn't

a good bargain, compared to other, more dense animals with heavier muscles.

Compare a grown adult man's 125,800 calories to a woolly rhinoceros's 1,260,000 calories

or an mammoth's 3,600,000 calories.

What kind of meat to look for if you're trying to survive in the wild is a no-brainer.

1.

Humans In The Lab.

In case you thought cannibalism was a dying practice, relegated to only those stricken

with the extreme misfortune of starvation and barbarians from centuries and eons past,

you're dead wrong.

In a tweet, well-known evolutionary biologist Richard Dawkins recently asked a bizarre question:

"What if human meat is grown?

Could we overcome our taboo against cannibalism?"

The tweet also included a link to an article about lab-grown meat, also known as in vitro

meat or clean meat.Obviously, meat grown in a lab doesn't require the death of an animal,

only a few stem cells from a living specimen.

The same process, of course, could be utilized to create human meat.

So perhaps those of you who want to try cannibalism may someday be able to do so with the blessings

of modern science.

It is unlikely, however, that there would be much of a market for lab-grown human meat,

although there would inevitably be some, such as performance artists, who would want to

try it.

That's all.

Thank you for watching.

And don't forget to subscribe !

Yours, Eva.

For more infomation >> 10 Facts About Human Cannibalism From Modern Science - Duration: 12:17.

-------------------------------------------

wire cable coiling and winding /take up and paying off machine - Duration: 0:27.

Tel: +86 18136540512

wechat : +86 18136540512

whatsapp/skype:+86 15001830288 E-mail:shanghaixj123@yeah.net

For more infomation >> wire cable coiling and winding /take up and paying off machine - Duration: 0:27.

-------------------------------------------

1.8 Introduction to optimization - Gradient descent extensions - Duration: 9:59.

[MUSIC]

In this video, we'll discuss advanced optimization techniques that allow to

improve gradient descent methods.

As we remember from the previous video, if our function is difficult,

for example, it has elliptic level lines like on this graph?

Then gradient descent will oscillate, and it will take many iterations for

it to converge to the minimum.

To improve this, we can somehow change our gradient descent methods.

For example, take mini-batch gradient descent.

In this method, we take some random m examples from our training set,

we approximate the gradient based on these m examples, and

then we'll make a gradient step.

So in this video, we'll suppose that we have some way to approximate gradient.

Maybe by mini-batch, maybe by stochastic gradient descent with just one point,

or maybe we calculate full gradient.

That doesn't matter, we just have some approximation of gradient, and

we allot it by gt.

And in this video, we'll discuss how can we modify the way we make a gradient step,

how we add a gradient or antigradient to our previous approximation, w t-1.

So let's start with momentum method.

In this method, we maintain additional vector h at every iteration.

To calculate h in the step t, we take this vector h at the step t-1,

multiply by some coefficient alpha.

And add the gradient at the current iteration, gt,

multiplied by the learning rate, eta t.

So actually, ht is just a weighted sum of gradients from all previous iteration, and

from this iteration, also.

And then we just make a gradient step in the direction of -ht.

So what is the point of this method?

Suppose that we have some function, it'll make gradient descent, and maybe for

some coordinates of our parameter vector.

Gradients always have the same sign, so they lead us to the minimum.

And for some coordinates,

the sign of the gradient changes from iteration to iteration.

So, vector ht would be large for component where gradients have the same sign on

every iteration, and will make large steps by this coordinates.

And for coordinates that change sign,

they will just cancel each other and ht will be close to zero.

So, ht cancels some coordinates that lead to oscillation of gradients,

and help us to achieve better convergence.

Here, we should somehow choose parameter alpha, for example,

we could just set it to 0.9, and to somehow choose learning rate, eta t.

In practice, this momentum method indeed leads to faster convergence.

So for our last function, with elliptic level lines,

we'll see picture like this one.

So here, we don't oscillate that much as on the previous picture, and

we converge better.

There is also an extension of momentum method called Nesterov momentum.

So in simple momentum method, on every iteration,

we calculate the gradient at current point w t-1.

We take a gradient step from it by gt, and we then get our momentum, ht.

But since it's certain that we'll move in the direction of the momentum,

it will be more clever to, first, step in the direction of ht

to get some new approximation of parameter vector.

And then to calculate gradient at the new point, w t-1 plus ht.

So in this case, we'll get a better approximation on the next step.

Mathematically, we take the current point, w t-1,

we subtract alpha multiplied by h t-1, and calculate the gradient at this new point.

And then we add the gradient at this new point, multiplied by the learning rate,

to the h from the previous iteration, alpha multiplied by h t-1.

And in practice,

this method indeed leads to better convergence than momentum method.

So momentum method and Nesterov momentum method work better with difficult

functions with complex level sets.

But they still require to choose learning rate, and they're very sensitive to it.

So now we'll discuss some other optimization methods that try to choose

learning rate adaptively, so we don't have to choose it ourselves.

One of such methods is AdaGrad.

So let's remember how we make a gradient step, just for

one coordinate g of parameter vector w.

To make a gradient step, we take j-th component of parameter

vector from the previous iteration, is w t-1 j.

And we subtract j-th component of the gradient at the current point,

to find out the next parameter approximation, w t j.

In AdaGrad, we've obtained additional value for

each parameter from all our parameter vector, G.

So to calculate it, we take G t-1, G from the previous iteration.

We take j-th component of it, and

we just add a square of gradient at the current iteration.

So essentially, G is a sum of squares of gradients from all previous iterations,

and then we modify our gradient step.

We divide our learning rate, eta t,

by the square root of G t j, plus some small number, epsilon.

We add epsilon, just to make sure that we don't divide learning rate by zero.

So this AdaGrad method, it chooses learning rate adaptively.

Here, we can just set eta t, our learning rate, to some constant, for

example, 0.01, and don't remember about it at all.

So this doesn't need to somehow

wisely choose learning rate, but it also has some disadvantages.

Our auxiliary parameter, G, accumulates squares of gradient, and

at some step it can become too large.

We'll divide our learning rate, eta t, by a large number, and

gradient descent will stop building.

So to somehow overcome this, we need some other methods like AdaGrad.

By the way, AdaGrad has another advantage,

it chooses its own learning rate for each example.

So suppose that we are analyzing texts, and

each feature from our sample corresponds to one word.

So for some frequent word that we see in every document,

we have gradient updates on every step, and we'll make smaller steps.

And for some rare words that we can met only in one of thousand or

ten thousand documents, we'll make large updates,

because they are rare, we don't need them very often.

And we need to move faster in the direction of these words.

Another matter that can improve AdaGrad is RMSprop.

This method is very similar to AdaGrad, but here we take an exponentially weighted

average of squares of gradients on every step.

So here, to calculate G j at the step t, we take G j from the previous step,

t-1, we multiply it by some coefficient alpha.

And then we add the square of the gradient, G t j, at this iteration,

multiplied by 1- alpha.

And then we use this G to divide our learning rate, just like in AdaGrad.

So this method overcomes the problem of large sums of square gradients.

And here,

our learning rate depends only on last examples from our gradient descent method.

Let's start with RMSprop method and slightly modify it.

So in RMSprop, we maintained an additional variable,

and it will be augmented by v t j.

And is just an exponentially weighted sum of gradients from all iterations.

So here, to calculate v for j-th component at the step t,

we take v from the previous step, v j t-1, multiplied by some coefficient beta 2.

And we add square of gradient in the current iteration,

g t j squared, multiplied by 1- beta 2.

And we can notice, that this approximation, v has some bias towards

zero, especially at first steps, because we initialize it with zero.

So to overcome this, we just divide it by 1- beta 2, in the degree of t.

So this normalization allows us to get rid of this bias.

At first steps, this normalization is large, and for

large t's, this normalization almost equals to 1.

And then we use this v to divide our learning rate, eta t.

But as we go from momentum method,

an approximation of gradient from one step can be noisy and lead to oscillations.

So let's smooth our gradients.

To do it, we maintain another auxiliary variable, m,

that is essentially a sum of gradients.

Not squares of gradients, like in v, but just gradients.

And it's calculated in the same way,

it's just an exponentially weighted average with coefficient beta 1.

And then we replace g, the gradient approximation in our gradient step,

by m, by weighted average of gradients from all previous iterations.

So this method, Adam, combines both momentum methods and

adaptive learning rate methods.

And in practice, it achieves better convergence and faster convergence.

In this video, we discussed some advanced optimization

techniques that can improve gradient decent.

For example, momentum method and Nesterov momentum method,

that allows us to work with some complex functions with complex level sets.

And AdaGrad, RMSprop method, that have adaptive learning rates, so

we don't have to choose eta t manually.

And there is also an Adam method, that combines both of this methods.

[MUSIC]